Ever since facts have been replaced by ideologies, the obvious has increasingly become a mystery.

Absolutely. Trust isn’t binary, and it isn’t unique to Chinese models either. Every (frontier) model reflects the incentives, values and constraints of whoever built it. That’s exactly why competition matters.

Unfortunately, based on many years of experience, I have to agree. However, I also see a light at the end of the tunnel, particularly in Europe, where there is a growing desire to break free from the US SaaS stranglehold, using open source software.

Regarding OWAI I see a fundamental difference: Slack isn’t just software, it’s a hosted service with identity, storage and network effects.

An open weight model is more like a compiler: once downloaded, nobody can revoke it. You may still pay for inference, but pricing power drops when anyone can host the same model.

Funny how »open is dangerous« only became the dominant narrative once Chinese models caught up: If your competitive moat depends on governments blocking free alternatives instead of building better products, maybe your moat isn’t technology at all.

Exactly. It’s much easier to regulate an API than software people can download and run on their own hardware. Once Since capable open weight AI models exist, they spread much like Linux or other open source software: mirrors appear, forks emerge, and improvements compound. Governments can make access less convenient, but preventing global distribution altogether is a far harder problem than restricting a hosted service.

AI companies keep promising productivity gains, but one of the first guaranteed outcomes is that everyone else helps pay the electricity bill. If the grid is upgraded for a few hyperscalers while the costs are spread across millions of customers, that’s not a free market. It’s a subsidy with extra steps.

The weird part is that everyone keeps calling it public infrastructure, but if the primary beneficiary is a handful of private AI companies, that argument starts looking pretty thin. Reliable grids benefit everyone. Dedicated power corridors for hyperscalers are a much harder sell, especially when locals bear the costs.

The awkward part is that software has a habit of racing toward free once it becomes good enough. If an open weight model delivers 95 percent of the value without recurring API costs, plenty of companies will choose that and spend the savings on integration instead. History keeps rhyming, even if investors hate the tune.

Have you finished school?

Are you still interested in Closed AI?

That’s not necessary: as the Open Weight AI models from the US are rather poor, nobody wants them anyway. 🤷

And on top of that, it’s still saddled with XTwitter’s debts.

I’d argue the models and software are the real long term assets. GPUs depreciate like any other hardware, but a better training pipeline, inference stack, proprietary data, and a model with millions of paying users can survive multiple hardware generations. The chips are replaceable. The ecosystem and customer relationships are much harder to replicate.

Perhaps. The difference with regard to OWAI, however, is that US companies have no alternatives to Chinese suppliers.

Maybe, but »excessive investment« and »failed technology« aren’t the same thing: Railroads, fiber optics and the dot com era all burned absurd amounts of capital, yet the infrastructure outlived the investors. AI could follow the same pattern: terrible returns for today’s shareholders, enormous value for tomorrow’s economy.

Using security as a convenient tool to sideline foreign competition is a slippery slope, especially from companies that built their own success on open research.

Funny how every dip was supposed to be a once in a lifetime buying opportunity until people actually bought near the top. SpaceX is still an incredible company, but incredible companies can still be overpriced. Reality eventually sends the invoice, even when the CEO is treated like a prophet.

That assumes demand is fixed. Historically, the biggest technology shifts created entirely new markets that barely existed beforehand. Almost nobody predicted today’s cloud or app economy from early internet revenue. AI spending could still prove excessive, but current demand is a poor ceiling for what future demand might become.

I think that’s probably the most underrated use case. We keep treating LLMs like faster search engines, but their real value may be surfacing relationships humans never think to test.

Wall Street spent years demanding AI investment, then panicked when AI investment showed up on the balance sheet: If Alphabet keeps printing record profits while building the infrastructure for the next decade, this may end up looking more like impatience than prudence. The market loves growth, until it has to pay for it.